National Air Cargo Boeing 747 Crashes Shortly After Takeoff

[FSTargetDrone]

This accident in this video (which evidently happened this past Monday, 29 April in Afghanistan) may be the result of cargo shifting inside the aircraft. It's quite chilling and disturbing though not graphic in terms of showing fatalities. Fair warning if you do not want to see this regardless.

There isn't much sound in the video other than the driver's cursing to himself but at about halfway through, a dog inside the vehicle with the driver starts to whimper loudly. So don't be startled by that if you have your sound turned up.

Story from NPR:

Video May Show 747 Jet's Last Seconds Before Bagram Crash

by Bil Chappell

April 30, 2013 5:39 PM

The final seconds of an American 747 civilian cargo jet's disastrous takeoff from Bagram Airfield in Afghanistan Monday — resulting in a crash that killed all seven crewmembers — were apparently captured by a dashboard camera. In the video, the aircraft is seen in a steep climb, until it fails to gain altitude and plummets into the earth.

The dashcam video shows the cargo jet, reportedly loaded with military vehicles, slowing in its ascent before making an excruciating turn toward the ground, where it burst into a fireball. The footage was posted at the website, which then had trouble with its servers. A copy was posted on YouTube. Warning: Some viewers may find the images disturbing.



Multiple aviation websites have also posted the video, including The Aviation Herald, which included this account:

"Several observers on the ground reported the National Air Cargo Boeing 747-400 had just lifted off and was climbing through approximately 1200 feet when it's nose sharply rose, the aircraft appeared to have stalled and came down erupting in a blaze... According to a listener on frequency the crew reported the aircraft stalled due to a possible load shift."

Others have suggested that a sudden powerful blast of wind may have disrupted the plane's climb.

In the video, the driver of the vehicle with the dashcam slows and reverses his direction as he sees the plane's nose turn back toward the earth. After the initial blast, he drives toward the crash site across open ground and around fencing, where thick black smoke rises from the wreckage.

At this point, we cannot say with absolute certainty that the haunting video is genuine. But the images agree with photographs and descriptions of the incident. In particular, a commenter named Scott wrote on the Loadstar website that he had witnessed the crash, saying that the plane "looked like it had flattened out to nearly level but had very little or no forward speed" at the point of impact. His description matches the footage.

The dashcam video includes a time stamp of 2013/02/01 — but that could have many explanations, including a lack of continuous power to the unit.

The AP has more information about the doomed flight, and its crew:

"The plane — owned by National Airlines, an Orlando, Florida-based subsidiary of National Air Cargo — was carrying vehicles and other cargo, according to National Air Cargo Vice President Shirley Kaufman. She said those killed were four pilots, two mechanics and a load master, who was responsible for making sure that the weight and balance of the cargo is appropriate. Five of the seven fatalities were from Michigan, said Kaufman."

In a statement, National Airlines said it will work with the U.S. National Transportation Safety Board and Afghan officials to determine why the jet crashed.

"This is a devastating loss for our family and we'll work diligently with authorities to find the cause," said National Airlines President Glen Joerger. "Most importantly, our thoughts and prayers are with our crewmembers and their families."

[Sea Skimmer]

The load was apparently five Strykers, 20 ton objects apiece. If one or more broke free in that situation, its all over. But such a failure would require some really glaring problems with either the plane, or both the loading crew and the aircrew who should have inspected the tiedown work. This isn't something which can happen from one chain breaking.

[Hawkwings]

Looking at the video, by the time the plane comes into view it's already at an extremely high nose-up attitude, so we don't really see what the beginning of it looked like. From what we do see in the video, the plane is trying to climb at a too-high angle, loses speed, the right wing stalls first, and the plane drops like a rock. It is not flying at any point after that, but simply falling. The pilot manages to level out the plane, but there is no forward velocity - certainly not enough to generate the lift needed for level flight.

It is certainly possible that one of the vehicles broke free and shifted to the back, causing the center of gravity of the plane to move too far back. That would cause an immediate nose-up sort of motion. What is puzzling to me is the length of time that the pane was attempting to fly in such a nose-up condition. I know they're probably trying to do a steep climbout, but to me it looks like as soon as the plane comes into view it is already flying far too nose up and far too slowly for things to be alright. Honestly it might not even be because of improper loading or weight shifting. This might simply be the pilot attempting to fly too far nose up and not realizing his airspeed was too low until too late. Which is extremely implausible, but so is the payload breaking loose.

[Stark]

I guess this means when transporting vehicles the back of a 747 is strong enough to prevent the cargo from just falling out? Does it mean the pilot/computer reacted slowly or that the control surfaces don't have the authority to offset such a change in COG?

[Grumman]

I guess this means when transporting vehicles the back of a 747 is strong enough to prevent the cargo from just falling out?

The later 747s can apparently carry more than 600 people, so five 20 ton vehicles distributed along the length of the fuselage probably isn't too different to 5*120 people distributed along the length of the fuselage, as long as they aren't moving about.

[Hawkwings]

First of all, I am not convinced that it was a shift in cargo that caused this. If we assume that is the case though, there are weight and balance diagrams for loading an airplane that stays within an acceptable CG range. For improved takeoff and climb performance, they may have loaded the plane to a more aft CG than "standard" (this is not uncommon though in my understanding). If this is the case, then perhaps the cargo did not break loose or anything but simply slid back a little - enough to cross the boundary into unsafe CG range. The rear CG limit is set by stability of the airplane and tail control authority. Since the 747-400 is not a fly-by-wire plane (meaning the pilot's controls go into a computer which then outputs control surface deflections) but rather old fashioned cables to the control surfaces, this most likely meant that the pilot was having to use his skills and reflexes and judgement to control an unstable plane. If it were smaller or flying faster then the crash may have happened faster. As it is, I believe a 747 at low speed is about at the limit of what a human pilot with human reflexes can fly in an unstable condition. So the pilot was probably working very hard keeping the plane under control early on in the video, but lost too much speed and stalled the right wing.

It is also entirely possible that the plane did not have enough elevator authority to pitch down, because of both the too far aft CG as well as the very low airspeed.

[Stark]

The later 747s can apparently carry more than 600 people, so five 20 ton vehicles distributed along the length of the fuselage probably isn't too different to 5*120 people distributed along the length of the fuselage, as long as they aren't moving about.

Right, but if an armoured vehicle on the loadbed was moving freely (and the plane was pitched up 60 degrees like in the video), I would have expected it to have moved to the back of the loadbed. The passenger 747s I've been in probably wouldn't handle an armoured vehicle punching a hole in their ass. It'd make sense if the cargo ones were built differently, of course.

Hah, I hadn't even considered that it'd be a plane without active flight control.

[Hawkwings]

Boeing planes and Airbus planes have a major philosophical difference in that Boeing believes the pilot should ultimately be in charge of the plane, whereas Airbus thinks that the computer should be in charge of the plane. Airbus pilots fly computers, and Boeing pilots fly airplanes.

Cargo 747s are not that different from passenger 747s. At least, not different enough that a big heavy armored vehicle would punch through one and not the other. The skin and ribs are the same, but the floor is stronger on a cargo 747, and of course the windows and doors are different as well. If a vehicle had indeed broken loose and slammed into the back of the plane, I would have expected to see damage and deformation of the rear fuselage.

[lPeregrine]

Does it mean the pilot/computer reacted slowly or that the control surfaces don't have the authority to offset such a change in COG?

Either, depending on the exact situation. If the center of gravity gets too far back the plane becomes unstable and once you cross a certain pitch angle* you can no longer lower the nose, but it could have also been pilot error in reacting too slowly to a smaller balance problem that wouldn't have been fatal with a faster reaction. Or it could have been various other problems: idiot pilot trying to climb too steep (someone elsewhere commented that you need a steep climb to clear terrain, and everyone wants to get above potential enemy fire asap) with a heavy load and high density altitude, mechanical failure resulting in sudden excessive up elevator, etc.


*This makes it more dangerous than COG too far forwards. If it's too far forwards you can't raise the nose, which at least makes it hard to get off the runway and die. But with the COG too far back the plane can fly just fine, other than needing constant pitch control inputs to remain level, up until you cross the line where you suffer a fatal loss of control somewhere far above the ground. If someone really screwed up the plane could have been loaded with a fatal balance problem and nobody would have noticed until it was too late.

[Hawkwings]

A fairly basic thing to note is that yes, a plane can fly just fine with a CG that is "too far" aft. You just need a computer flight control system constantly making small adjustments in the control surfaces to keep you straight and level. This is how fighter jets get their great maneuverability: they fly around unstable and rely on the computer to keep it in their air. If you took a modern fighter jet and turned off the computer, it would be tumbling out of control in a few seconds. The pilot wouldn't even have time to react because the forces on the plane are large and the plane's inertia is small. However, if you have a slow heavy plane that is unstable, the instability-induced pitching and whatnot is slow enough that a skilled pilot can control it. The Wright Flyer was not stable, but it was controllable.

[Sea Skimmer]

Its not like a load shifting Stryker is a needle like impact on the rear pressure bulkhead. Those things are physically huge, and plainly the shift had to occur long before the aircraft was 60+ degrees nose up. I can easily believe one would shift and not punch out the back of the plane. Big planes are massively strong, the only reason they seem weak is the strength tends to be demonstrated in impacts with the solid ground.

[Beowulf]

Boeing planes and Airbus planes have a major philosophical difference in that Boeing believes the pilot should ultimately be in charge of the plane, whereas Airbus thinks that the computer should be in charge of the plane. Airbus pilots fly computers, and Boeing pilots fly airplanes.

Cargo 747s are not that different from passenger 747s. At least, not different enough that a big heavy armored vehicle would punch through one and not the other. The skin and ribs are the same, but the floor is stronger on a cargo 747, and of course the windows and doors are different as well. If a vehicle had indeed broken loose and slammed into the back of the plane, I would have expected to see damage and deformation of the rear fuselage.

This 747 was formerly a passenger aircraft. It was converted to cargo configuration.

[Sea Skimmer]

A number of 747s were produced as hybrids, with reinforcement to be freighters and the opening nose, but also all the kit and windows to carry a passenger deck. If this was a conversion it would have had to be one of those I think. IIRC they don't have the big rear side hatch of the dedicated freighter.

[TimothyC]

A number of 747s were produced as hybrids, with reinforcement to be freighters and the opening nose, but also all the kit and windows to carry a passenger deck. If this was a conversion it would have had to be one of those I think. IIRC they don't have the big rear side hatch of the dedicated freighter.

After doing some quick research, this was one of the conversions, but it didn't have the nose, only the aft hatch.

[Sea Skimmer]

Hmm, never knew those existed. That certainly throws all claims of the aircraft transporting armored vehicles out the window then.

[TimothyC]

Hmm, never knew those existed. That certainly throws all claims of the aircraft transporting armored vehicles out the window then.

Not really. The hatch is certainly tall enough* (It's just over three meters tall while a striker is not even 2.7m). You've have to angle it in, and I'm not saying it would be easy, but it looks like the geometry doesn't entirely preclude it. You might even have a harder time sliding one in via the nose because the forward cargo area isn't full height thanks to the deck above it.

*Source: Boeing Airport Planing Characteristics PDF

[CaptHawkeye]

It is also entirely possible that the plane did not have enough elevator authority to pitch down, because of both the too far aft CG as well as the very low airspeed.

It apparently didn't have enough power in time either. If you listen closely to the video, I believe you can hear the engines spooling up to full thrust. You're normally not supposed to run turbine engines at their actual full power but in a situation like this i'm sure the pilots cranked all 4 engines up as far as they could go. Turbines need time to spool up though and even if they were at full power, with a full load and takeoff fuel they would never have developed enough power to get to a controllable speed before the stall developed. Not at that pitch angle.

The right wing stalled almost right as you hear the engines delivering more power. The plane leveled out near the end but they were most definitely in an accelerated stall anyway which would not at all improve their situation.

This is just why CG planning and weight management on airplanes is such a serious thing. Light airplanes crash all the time because of poor weight and balance, though light planes tend to be more susceptible to overloading rather than operating outside the CG range. I'm interested to hear where the mistakes were made in this accident. Planning, piloting, management, or all of the above?

[Sea Skimmer]

Not really. The hatch is certainly tall enough* (It's just over three meters tall while a striker is not even 2.7m). You've have to angle it in, and I'm not saying it would be easy, but it looks like the geometry doesn't entirely preclude it.

I really can't believe they'd bother to do an operation that damn dangerous and time consuming that high off the ground when a basically unlimited supply of An-124 and Il-76 charter flights exist, and ramp space at Bagram, while vast, is barely adequate for the massive requirements put on the base.

I suppose though angling in 4x4 MRAPs might be a little more reasonable, and one of those breaking free would still be very bad.

[Hawkwings]

Loading in cargo through the side door diagonally is hardly dangerous or time-consuming. How do you think every non-747 freighter does it?

[Zaune]

I really can't believe they'd bother to do an operation that damn dangerous and time consuming that high off the ground when a basically unlimited supply of An-124 and Il-76 charter flights exist, and ramp space at Bagram, while vast, is barely adequate for the massive requirements put on the base.

Don't bet on it. Can you imagine what the Republicans would have to say if a photo of US military stores being loaded onto a Russian-made and quite possibly Russian-owned aircraft showing up in Aviation Week?

[Sea Skimmer]

Loading in cargo through the side door diagonally is hardly dangerous or time-consuming. How do you think every non-747 freighter does it?

The same way the 747 normally does it. Load at a 90 degree angle, then turn inside. Instead of actually going through the door at an angle as would be required here. I've never seen cargo handling gear that could even mate up to the plane like that.

Don't bet on it. Can you imagine what the Republicans would have to say if a photo of US military stores being loaded onto a Russian-made and quite possibly Russian-owned aircraft showing up in Aviation Week?

No. NATO actually has a number of Russian owned and crewed An-124s on long term lease for Afghanistan, and charters and leases of additional An-124 and Il-76 air frames are constantly in motion alongside lots of other civilian airlift for NATO and US specific use. Only Russia has a serious number of widebody ramp equipped airlifters available for charter in the world, and only Russian airlines fly the An-124 at all. So they get used because its far better then having huge amounts of arms, ammunition and armored vehicles stolen or destroyed in Pakistan, the northern route does not allow combat material to be shipped at all.

http://kf4lmt.files.wordpress.com/2012/ ... resize.jpg
Hey look a picture of a US MRAP being loaded onto a Russian airlifter!


This ones cool too and small enough to inline, damaged Chinook coming out of another An-124.

Use of chartered Russian airlift is basically essential to maintaining the war in Afghanistan as it has been fought.

[Hawkwings]

Loading in cargo through the side door diagonally is hardly dangerous or time-consuming. How do you think every non-747 freighter does it?

The same way the 747 normally does it. Load at a 90 degree angle, then turn inside. Instead of actually going through the door at an angle as would be required here. I've never seen cargo handling gear that could even mate up to the plane like that.

http://www.boeing.com/commercial/airpor ... ochure.pdf

Page 9 of this presentation shows several cargo arrangements for the 777 freighter, which only has the side door. Notice that it can fit 96" x 238.5" pallets, which must be loaded diagonally. It is not such a great leap of logic to assume that there must be cargo loading equipment that can load diagonally into a freighter.

[Surlethe]

What is puzzling to me is the length of time that the pane was attempting to fly in such a nose-up condition.

I have a couple of heuristic reservations about this conclusion just based the fact that a 747 is enormous. First, at the start of the video, it's far enough away that I'm not sure we can trust intuitive judgment on its angle or speed. Second, even a fast adjustment in its angle of attack, because of its size, will take time, perhaps seconds. But you seem to know a lot more about these behemoths than I do, so I'll just submit this and wait to hear what you say.

[Broomstick]

It is certainly possible that one of the vehicles broke free and shifted to the back, causing the center of gravity of the plane to move too far back. That would cause an immediate nose-up sort of motion. What is puzzling to me is the length of time that the pane was attempting to fly in such a nose-up condition. I know they're probably trying to do a steep climbout, but to me it looks like as soon as the plane comes into view it is already flying far too nose up and far too slowly for things to be alright. Honestly it might not even be because of improper loading or weight shifting. This might simply be the pilot attempting to fly too far nose up and not realizing his airspeed was too low until too late. Which is extremely implausible, but so is the payload breaking loose.

Payloads have been known to break loose before. It's not supposed to happen but it can and sometimes it does. It could be either human error or a snapped tie-down or both. We may never known.

At full power an airplane can continue to fly in a surprisingly nose-high attitude. The average person never sees what airplanes are actually capable of doing because normal practice is to get nowhere near the edges of the flight envelope. In combat and hazardous areas the SOP is to use best climb-out configuration to get up and away from possible missile attacks so they're already pretty nose-high.

Keep in mind, too, that the angle from which the video is shot will tend to emphasize the underside of the airplane, making it look even more nose-high than it actually is. Reports that the nose was 60 degrees above the horizon are certainly exaggerated, the result of optical illusions due to the angle of observation, or due to the inexperience of the viewer. A 747's wing stalls at an angle of attack of about 17-18 degrees. Full power + upward momentum will allow greater than 18 degree with respect to the horizon, but not enough to allow 60 degrees (unless it's going into a tail slide, which this one didn't). It's not an airshow stunt plane with an oversized engine that can climb on pure thrust.

I guess this means when transporting vehicles the back of a 747 is strong enough to prevent the cargo from just falling out?

Pretty much. The cargo slamming into the back of the fuselage may certainly cause some damage but it may not be apparent from the outside.

Does it mean the pilot/computer reacted slowly or that the control surfaces don't have the authority to offset such a change in COG?

Past a certain point the control surfaces can't compensate for a too-far aft COG. Slow reactions can also factor into the matter, but if it did result from cargo-shifting the situation may simply have been unrecoverable.

First of all, I am not convinced that it was a shift in cargo that caused this. If we assume that is the case though, there are weight and balance diagrams for loading an airplane that stays within an acceptable CG range. For improved takeoff and climb performance, they may have loaded the plane to a more aft CG than "standard" (this is not uncommon though in my understanding). If this is the case, then perhaps the cargo did not break loose or anything but simply slid back a little - enough to cross the boundary into unsafe CG range.

No sane pilot/cargo master is going to load the cargo to the very edge of the CG limits. A small shift in position isn't enough to throw the airplane into an unrecoverable position.

The rear CG limit is set by stability of the airplane and tail control authority. Since the 747-400 is not a fly-by-wire plane (meaning the pilot's controls go into a computer which then outputs control surface deflections) but rather old fashioned cables to the control surfaces, this most likely meant that the pilot was having to use his skills and reflexes and judgement to control an unstable plane.

Actually, no, 747's do not use “old fashioned cables” from the controls to the flight surfaces. There are a bunch of hydraulics in between, plus various systems to assist the pilot in flying.

I would also like to point out that past a certain change in attitude ALL autopilots shut down, relying on the human up front to fly the airplane. The black boxes from the Air France Flight 447 crash clearing show where the airplane turns over authority to the humans (later, when I don't have to rush to get to work, I can provide a better link/quote). Why does this happen? Because while computers are excellent with routine matters humans are still better at handling the unexpected. If the same thing had happened in a Airbus cargo plane the flight computer would be relying on human decision-making.

[Broomstick]

A fairly basic thing to note is that yes, a plane can fly just fine with a CG that is "too far" aft.

No, it can't. Place the CG too far aft and the flight controls can't compensate regardless of whether there is a human or a computer in control. That's one of the reasons shifting cargo can be dangerous, even in something like an Airbus. And Airbus, in general, won't let the pilot make a fatal error. If you attempt a maneuver that could be lethal then flight computer simply won't obey that command, or will only obey within safe parameters. The problem is that shifting cargo is not within the control of the flight computer and can result in a CG so aft as to be fatal.

You just need a computer flight control system constantly making small adjustments in the control surfaces to keep you straight and level. This is how fighter jets get their great maneuverability: they fly around unstable and rely on the computer to keep it in their air.

You are mistaking “inherent instability” for “out of envelope CG” and they are not the same thing. Fighter jets also have documented CG limits and if you exceed them no flight computer on (or off) Earth is going to save your ass. Of course, as most fighter jets are single seat and don't normally carry variable loads being out of CG is vanishingly rare. CG limits are most tricky in passenger and cargo planes where the payload varies significantly from one flight to the next. It can also be tricky in ultra-long-duration flights like the nine-day non-stop flight of the Voyager where the total fuel at takeoff exceeded the weight of everything else – fuselage, people, etc. - combined and part of fuel management was burning off/shuttling fuel between the various tanks to keep everything properly balanced out.

The Wright Flyer was not stable, but it was controllable.

It was not, however, out of CoG limits because the flight controls were (just barely) adequate to maintain control.

It apparently didn't have enough power in time either. If you listen closely to the video, I believe you can hear the engines spooling up to full thrust. You're normally not supposed to run turbine engines at their actual full power but in a situation like this i'm sure the pilots cranked all 4 engines up as far as they could go. Turbines need time to spool up though and even if they were at full power, with a full load and takeoff fuel they would never have developed enough power to get to a controllable speed before the stall developed. Not at that pitch angle.

The right wing stalled almost right as you hear the engines delivering more power. The plane leveled out near the end but they were most definitely in an accelerated stall anyway which would not at all improve their situation.

The truly fatal thing here is that they ran out of altitude. From speaking with airline pilots who flew back in the days before computer-controlled flight simulators, back when pilots took real airplanes like the 747's and did real stall recovery practice, even under the best circumstances a jet like that lose a hell of a lot of altitude during even a well-managed stall, enough that allowing for a minimum drop of 10,000 feet was standard practice (about 3,000 meters for the metric speakers). While it is possible for the altitude loss to be less than that as a practical, real-world matter you need to allow that much room. If you stall a 747 with less than that under your wings you stand a very high probability of dying. They were simply too low, far too low, to survive a stall in that airplane.

If they had been much, much higher they might have survived. I base that on the fact that, instead of corkscrewing into the ground nose-first the pilot(s) arrested the incipient spin and were leveling out the airplane. A few more thousand feet and gravity power would have provided enough airspeed over the wings to get the airplane flying again. Maybe. If nothing else went wrong. Maybe when the nose dropped lose cargo slid forward again, putting the CG back into survivable limits. I'm getting way into speculation here, but it's possible.

I'm interested to hear where the mistakes were made in this accident. Planning, piloting, management, or all of the above?

My guess is likely a bit from all columns, which is how accidents usually are these days. You have to have multiple failures in the redundant systems (by which I mean SOPs and human actions as well as mechanical and computer systems) for an accident to occur.

I was hoping to find the exact term used in Airbus-speak for when the computer essentially says “I give trying to understand this, here, you humans make decisions” but my google-fu is weak today. Anyone I referenced it in one of the Air France 447 threads. If I can find it I'll give you all the correct term.